REVIEW ARTICLE

Understanding the Bacterial Flora of the Female

Genital Tract
Bryan Larsen1 and Gilles R. G. Monif2
1

Des Moines UniversityOsteopathic Medical Center, Des Moines; and 2Creighton University School of Medicine, Omaha

QUALITATIVE MICROBIOLOGY
The microbiological flora of the lower female genital
tract is a dynamic, complex example of microbial colonization, the regulation of which is not fully understood. Much of what we know about the bacterial composition of the female genital tract is derived from
qualitative, descriptive studies [110]. The fund of information that such studies have provided with regard
to the microbial flora of the lower female genital tract
is weakened by the intrinsic technical limitations that
are inherent in the studies. Often, even the usefulness
of qualitative data is negatively affected by inappropriate or suboptimal methods of data collection, failure
to use appropriate transport systems or enriched media,
or a lack of stringent anaerobic technique in the processing and culture of specimens.
The importance of using specialized media is illustrated in a study of Clostridium difficile by Bramley et

Received 31 March 2000; revised 26 June 2000; electronically published 9

February 2001.
Correspondence: Dr. Bryan Larsen, Office of the Dean for University Research,
Des Moines UniversityOsteopathic Medical Center, 3200 Grand Ave., Des Moines,
IA 50312-4198 (bryan.Larsen@dsmu.edu).

al. [11]. These investigators evaluated cultures of vaginal specimens obtained from 522 women who made
a total of 902 visits to a family planning clinic, and they
found this organism in only 1 patient. However, when
a specialized medium that contained 0.2% para cresol
was used, a higher rate of isolation (11%) was obtained.
One can only speculate as to how many more microbial species would have been recovered if truly optimal media and methods had been used for all studies
reported in the literature prior to the 1970s. Isolation
techniques used prior to the 1970s resulted in a gross
underestimation of the importance of anaerobic bacteria as major constituents of the normal flora of the
female genital tract. Failure to use appropriate transport
systems as well as failure to use optimal media and
anaerobic culture techniques have compromised the results of many studies with regard to the delineation of
the bacterial constituents present.
Although anaerobic bacteria had been identified previously, it was not until publication of the work of
Gorbach et al. [2] and, soon afterward, that of Ohm
and Galask [3], Galask et al. [5], and Hill [7], that the
role of anaerobic bacteria both in maintaining health

Clinical Infectious Diseases 2001; 32:e6977

Q 2001 by the Infectious Diseases Society of America. All rights reserved.
1058-4838/2001/3204-00E1$03.00

Flora of the Female Genital Tract

CID 2001:32 (15 February) e69

Downloaded from http://cid.oxfordjournals.org/ by guest on August 28, 2015

The microbiological flora of the lower female genital tract provides a dynamic, complex example of microbial
colonization, the regulation of which is not fully understood. When an exogenous bacterial species, with its
array of virulence factors, is introduced into the host, disease does not always occur. Conversely, under selected
conditions, commensal endogenous bacteriafor example, Gardnerella vaginalis and group B streptococcican participate in disease processes. Disease caused by both exogenous and endogenous bacteria correlates positively with a markedly increased level of bacterial replication. The key question is what determines
the quantity of a given bacterium at any given time. For disease to occur, exogenous or endogenous bacteria
that possess pathogenic prerequisites must attain replicative dominance. Their ability to do so is potentially
governed by inhibitory or synergistic interrelationships with other microbes.

and in causing disease became more clearly defined. Gorbach

et al. [2] demonstrated that, in women of reproductive age,
anaerobic bacteria outnumbered aerobic bacteria in a ratio of
approximately 10:1. This ratio clearly reflects a dynamic colonization process. For example, although adolescent subjects
appeared to have a greater prevalence of anaerobic bacteria,
aerobic bacteria appeared to become more abundant with advancing age, onset of sexual activity, and parity. A study of
postmenopausal women who were either receiving or not receiving estrogen replacement therapy found that such therapy
had no effect on facultative organisms; however, anaerobic isolates tended to be less prevalent among women who received
such therapy. A notable exception, however, were anaerobic
lactobacilli, which appeared to be more prevalent in the tissue
of women receiving estrogen therapy [12].

Combined qualitative and quantitative studies require a

quantum increase in technical effort and, as a consequence,
tend to be limited in scope despite yielding richer information
[1318]. Recent studies have begun to focus more on the fact
that the density of microbial colonization appears to be relevant
not only to the condition of asymptomatic individuals but,
also, to the initiation of disease states, in which it is a critical
factor [13, 18, 19]. The microbial load for a given organism
appears to influence the relative risk of symptomatic infection;
however, in the absence of quantitative data, data that have
been extrapolated from qualitative studies (e.g., the prevalence
rates of individual species) are used as a surrogate for quantitative data. The concept exploited is that organisms of which
there are a great number are readily found in cultures, whereas
those species that are fewer in number may not be noticed
during primary isolation.
Quantitative studies of upper and lower female genital tract
disease due to exogenous bacterial species (e.g., Neisseria gonorrhoeae) and endogenous bacterial species (e.g., Gardnerella
vaginalis) have demonstrated one common finding: increased
numbers of bacteria are found during the course of disease.
The studies that have been published to date, although technically imperfect, do provide some information regarding the
dynamics of the bacterial flora of the female genital tract.

BACTERIOLOGICAL STUDIES OF THE NORMAL

FLORA
Studies of the normal bacterial flora of the female genital
tract are primarily limited to characterization of the types of
bacteria present in women who do not have identifiable disease.
Studies by Bartlett et al. [5], Larsen and Galask [8], and Gop-

e70 CID 2001:32 (15 February) Larsen and Monif

PATHOGENS AND COMMENSAL ORGANISMS

Within colonized tissues, such as those of the female genital
tract, what constitutes a pathogen is dependent not only on
the type of offending microorganism and its intrinsic virulence
but, also, on the species complexity of the florathat is, the
relative dominance, in numbers, of the various bacteria that
can be recoveredin individual asymptomatic patients. According to traditional thinking, a pathogen was a microbe that
was genetically endowed with a factor that, when expressed,
caused disease. This postulate became central to the concept
of the monomicrobial etiology of infectious diseases, which was
derived from correlation of the disease back to the etiological
agent. Examples that fit this concept well are diseases caused
by N. gonorrhoeae or Treponema pallidum.
However, the mere presence of an unknown, exogenous,
potentially pathogenic species does not necessarily constitute
disease when disease is defined in terms of symptoms. Understanding how specific bacteria produce disease has been tied
Table 1. Prevalence of aerobic (facultative) isolates reported
in vaginal flora studies published in the literature.
Prevalence in
vaginal flora, %
Aerobic isolate

Low

Mean

High

Gram-positive rods
Diphtheroids
Lactobacilli

40

80

18

60

90

Gram-positive cocci

Staphylococcus aureus

25

Staphylococcus epidermidis

50

95

Streptococcus species
a-Hemolytic

20

38

b-Hemolytic

15

22

Nonhemolytic

20

32

Group D

28

45

Escherichia coli

18

33

Klebsiella and Enterobacter species

10

20

Proteus species

10

Pseudomonas species

0.1

Gram-negative rods

NOTE.

Design of table is based on [83].

Downloaded from http://cid.oxfordjournals.org/ by guest on August 28, 2015

QUANTITATIVE MICROBIOLOGY

plerud et al. [12] (see tables 1 and 2) have effectively delineated

the principal bacteria that reside in the female genital tract,
although they have not delineated their quantitative interrelationship. In terms of planning empirical therapy, it may be
just as important to know which organisms are not isolated
with high frequency as it is to know which organisms are commonly isolated.

Table 2. Prevalence of anaerobic microorganisms present

in cultures of cervical and vaginal specimens obtained from
asymptomatic women (according to the results of selected
reports).
Percentage of isolates,
according to reference
Organism

[20]

[21]

[22]

[23]

[4]

B. biviusa

12

B. fragilis

17

40

12

16

Bacteroides species

B. melaninogenicusb

33

Other

40

18

46

10

Bifidobacterium species
Clostridium species
C. perfringens

13

Any

Eubacterium species

31

Fusobacterium species

28

13

Gaffkya speciesc

Lactobacillus species

10

46

52

P. asaccharolyticus

48

12

P. magnus

11

17

P. prevotii

17

21

Other

11

33

64

65

P. anaerobius

34

15

P. intermedius

10

P. micros

Peptococcus speciesd

Any

Peptostreptococcus species

P. productus

Any

33

76

35

15

Proprionibacterium species

Veillonella species

27

11

NOTE.
a
b
c
d

Dashes denote that no specific information is available.

Prevotella bivia.
Prevotella melaninogenica.
Aerococcus species.
Peptostreptococcus species.

to knowledge of virulence properties, which allow the bacteria

to function as monoetiological agents. Such microorganisms as
Neisseria gonorrhoeae, Streptococcus pyogenes, Streptococcus
pneumoniae, Haemophilus influenzae, Listeria monocytogenes,
and Trichomonas vaginalis are not ordinarily part of the flora
of the female genital tract. They bring the potential for disease
to the vaginal/endocervical area by virtue of their inherent biological properties, although the presence of these properties
clearly does not guarantee that disease will occur. Once the
normal bacterial constituents of the female genital tract are

Flora of the Female Genital Tract CID 2001:32 (15 February) e71

Downloaded from http://cid.oxfordjournals.org/ by guest on August 28, 2015

Other

defined, one is confronted with having to explain why apparently commensal bacteria (e.g., G. vaginalis, group B streptococci and Escherichia coli) can cause disease.
More than a century after Pasteur introduced the idea of
monoetiological disease (the idea that 1 pathogen causes 1 disease), we still struggle with the definition of the term pathogen [24]. In the middle of the 19th century, Pasteur provided
evidence that the presence of an organism, such as the anthrax
bacillus, in a host is associated with disease; however, there
frequently has been a tendency to think that the mere presence
of certain organisms is synonymous with disease. It was not
until the early part of the 20th century that Theobald Smith
introduced the idea that disease was the result of the interplay
between microbial virulence, dominance of the organism in
terms of numbers, and the innate defenses of the host [25].
Because Kochs postulates stated that monoetiological disease
could be demonstrated through production of infections in
animals, many studies were done in which animals could be
successfully infected with microbial pathogens. Often, however,
these infections eventuated only when a change in the microenvironment was created as part of the experimental model of
infection. For example, peritonitis might be more efficiently
induced when blood is added to the inoculum. Gangrene has
been known to develop when calcium chloride is implanted
into muscle along with the Clostridium species. In rodents,
vaginal infection with Candida albicans requires estrogen replacement therapy for the animal host. Bacteria that are normal
constituents of the vaginal flora of the host have the potential
to cause symptoms of disease, but they apparently require some
alteration in the microenvironment to do so. C. albicans, group
B Streptococcus (GBS), G. vaginalis, and Escherichia coli, which
are organisms that are commonly isolated from the lower female genital tract, can, under select circumstances, cause disease.
Although they are not indigenous to the microflora of the
genital tract, organisms that are commonly termed pathogens, such as H. influenzae, S. pneumoniae, S. pyogenes, and
T. vaginalis, may also be present without causing symptoms,
much in the way that organisms that are part of the normal
flora of the genital tract are typically present. These seeming
inconsistencies focus on a critical question regarding the pathogenesis of infectious diseases: what enables a given organism
to produce disease?
For endogenous bacteria of the female genital tract, the microbiological environment may supersede selected inherent virulence factors in terms of importance; at the very least, the
microbiological environment may affect the bacterial expression of virulence factors [26]. Theoretically, if a virulence factor
is constitutive, the number of organisms present will determine
the amount of the virulence factor available to promote infec-

e72 CID 2001:32 (15 February) Larsen and Monif

dance. Savage [35], in a classic study of the gastric flora of

mice, found that 2 distinct tissue areas exist, and these are
normally colonized by a nearly pure culture of Lactobacillus
species and a nearly pure culture of yeast, respectively. When
Lactobacillus species were eliminated by means of antibiotic
treatment, the yeast took over the microbiological void. This
observation suggested control of yeast through an antagonistic
effect of the Lactobacillus species. Conversely, Candida species
may exert their own antagonistic effects on populations of
bacteria.
Monif and Carson [36] compared the patterns of isolates
from women with and without co-isolation of C. albicans. In
the presence of C. albicans, selected aerobic and anaerobic
gram-positive isolates were diminished in a statistically lower
frequency. Hipp et al. [37] had previously demonstrated that
Candida species can produce a substance that suppresses the
growth of N. gonorrhoeae. Later, Shah and Larsen [38, 39]
showed that Candida species can produce gliotoxin, which,
along with other inhibitory compounds in appropriate concentrations, is antagonistic toward various bacteria.
G. vaginalis: bacterial vaginosis (BV). The relationship
of G. vaginalis to disease is not simply a phenomenon of cause
and effect. G. vaginalis can be a common constituent of the
vaginal flora of women [40, 41], yet only a relatively small
percentage of these women have symptoms or have a clinically
significant vaginal discharge. McCormack et al. [40] identified
the presence of G. vaginalis in the vaginal samples obtained
from 150 of 446 women who visited a student health center
and who were free of clinically overt disease. Again, the difference between colonization and disease appears to be partially
a function of the magnitude of replication of bacteria. Quantitative bacteriological studies have shown that symptoms that
involve G. vaginalis are associated with 1107 cfu per gram of
vaginal fluid [18]. If reintroduction occurs after therapy as a
result of sexual contact with an untreated partner, the patient
is usually asymptomatic; however, in such patients, the quantitative counts are !105 cfu per gram of vaginal fluid. For disease
to occur, not only must there be an environment that will
sustain G. vaginalis as a constituent of the microbiological flora,
but something must happen to free the bacteria from the inhibitory restraints that govern the magnitude of its replication.
The complexity of microbial interrelationships is further suggested by the finding that effective clinical and microbiological
cures can be achieved by use of metronidazole in only
75%80% of patients [41, 42]. For metronidazole to function
effectively, an organism must have a functional nitroreductase
system. Only 15%22% of Gardnerella isolates have this enzyme
system. This observation has led investigators to speculate that
the mechanism by which metronidazole has its effect involves
its impact on the concomitantly flourishing anaerobic bacteria

Downloaded from http://cid.oxfordjournals.org/ by guest on August 28, 2015

tion. The number of organisms may be controlled by means

of the antagonistic or synergetic interaction between the different microbial species present. Three relatively common conditions that involve the female genital tractnamely, vaginal
candidiasis, bacterial vaginosis, and infection with GBSshow
evidence of regulation.
C. albicans: vaginal candidiasis. During the past several
decades, the many published surveys of vaginal flora specimens
obtained from asymptomatic women have clearly shown that
C. albicans may be present without the typical symptoms of
yeast vaginitis. In a study by Glover and Larsen [27], the results
of successive cultures of vaginal flora specimens obtained from
women who were followed throughout pregnancy indicated
that Candida species may be present in stable association with
the genital epithelium. Moreover, the majority of women who
have vaginal yeast also carry the organism in the gut. The typical
rate of yeast carriage varies among populations and increases
both after puberty and during pregnancy, which suggests an
important role for host physiology in cases of vaginal candidiasis.
A relationship between estrogen levels and bacterial colonization has been recognized almost since the inception of
studies of normal vaginal flora; this relationship holds true for
Candida species as well. For example, rats are resistant to colonization by Candida species, unless the animals have an
amount of estrogen sufficient enough to cause vaginal cornification [28]. Growth of bacteria in the flora of the genital tract
is stimulated by estrogen [28, 29]. Prevalence of Lactobacillus
species and prevalence of yeast in different populations tend
to show that the times when prevalence of Lactobacillus species
is highest (during the reproductive years and, especially, during
pregnancy) are also the times when the prevalence of Candida
species is highest. Hydrogen peroxideproducing Lactobacillus
species may co-colonize with Candida species [28]. Although
Candida species are less susceptible to the microbicidal effects
of hydrogen peroxide than are noncatalase producers, such as
N. gonorrhoeae and Streptococcus agalactiae, Candida species
could be inhibited by hydrogen peroxide [30, 31]. This is presumably due to the fact that hydrogen peroxide damages cellular membranes unimpeded by the intracellular catalase [32].
Classically, vulvovaginal candidiasis occurs in association with
a significant increase in the number of colony-forming units
of Candida species that are present in the tissue-invasive form.
Any microbiological influence that allows the yeast concentration to increase may result in the development of symptoms,
and any microbiological effect that suppresses the number of
yeast could ensure that it remains as a commensal organism
[33, 34].
Both studies of animal models and observation of humans
suggest that there is an inverse relationship between bacterial
and yeast floras with respect to prevalence and numerical abun-

tions emphasize the importance of discovering what regulates

the number of colony-forming units per gram of vaginal fluid.
Chaisilwattana and Monif [57] have published the most extensive study that explores the ability of GBS to inhibit grampositive and gram-variable constituents of the bacterial flora of
the female genital tract. By use of an agar overlay assay technique, test strains of GBS were first inoculated and then were
allowed to reach a level of heavy growth. The plate was then
overlaid with new media. The target strain was then inoculated
onto the fresh agar and was incubated to achieve heavy growth.
The GBS test panels uniformly inhibited group A, B, C, and
G streptococci, lactobacilli, G. vaginalis, and most diphtheroid
strains. Variable inhibition by GBS was observed with viridans
streptococci, nonhemolytic (neither group B nor group D)
streptococci, peptostreptococci, and enterococci. The GBS test
panels did not inhibit the growth of either coagulase-negative
staphylococci or S. aureus. The 23 GBS isolates from neonates
or adults with septicemia did not differ from the 18 isolates
from subjects without septicemia, with regard to their ability
to inhibit the challenge bacteria. When converse testing was
done, the growth of GBS isolates was uniformly inhibited by
coagulase-negative staphylococci and by the majority of the
enterococci, but it was not inhibited by S. aureus.
The stab/chloroform technique was used for confirmation
of inhibition. In this technique, the GBS isolate is embedded
into the agar, which is then incubated for 24 h. The GBS colonies are then killed as a result of exposure to chloroform, and
the plate is then covered with the challenge bacteria. In addition
to confirming the results of the overlay agar assay, the stab/
chloroform technique demonstrated that, regardless of the
cause of the inhibition, the presence of live GBS was not a
prerequisite and that the inhibitor had the ability to diffuse
through the agar. Both in vitro techniques require that the
density of GBS reach maximal levels of growth.
Qualitative studies of cultures of vaginal/cervical bacteriological specimens obtained from asymptomatic women challenge the interpretation of in vitro observations. GBS are not
infrequently isolated along with lactobacilli, G. vaginalis, coagulase-negative staphylococci, and enterococci. Deeper analysis does reveal data that are compatible with those of in vitro
studies. In a large qualitative study, Carson et al. [58] demonstrated that whenever GBS was isolated, no other b-hemolytic Streptococcus species was present. Conversely, when another b-hemolytic Streptococcus species was present, GBS was
not concomitantly present.
The results of in vitro studies of bacterial inhibition may be
difficult to relate to the results of qualitative studies of vaginal
flora. The ability of GBS to inhibit bacterial growth in vitro is
based on it being present in a concentration of >109 cfu/mL,
and such high-density populations of GBS may not often be
seen in vivo. In a review of a number of clinical studies that
Flora of the Female Genital Tract CID 2001:32 (15 February) e73

Downloaded from http://cid.oxfordjournals.org/ by guest on August 28, 2015

population, which sustains dominance in conjunction with

Gardnerella species [43].
The complexity of interspecies interactions has not been
completely unraveled with regard to BV. Among the prevailing
conditions in the patient who has BV are an elevated pH level
(range, 55.5) and the presence of various primary amines and
polyamines, which can be detected on the basis of the presence
of the characteristic fishy smell that they emit after KOH has
been added [44, 45]. It is not clear what precipitates the condition, and several causes could be suggested. A decrease in the
number of lactobacilli could result in decreased production of
hydrogen peroxide and acid, thereby allowing for unrestricted
growth of other constituents of the flora. A shift in the redox
potential favoring anaerobic bacterial may in turn contribute
to proteolysis, thereby leading to a more alkaline environment
[46]. The proteins and amino acids that are released may be
metabolized by anaerobic bacteria as a result of the production
of amines, which contribute to both the alkalinization of the
vaginal environment and the odor problem, and which may
also contribute to vaginal irritation [47]. Some anaerobic bacteria produce succinic acid, which is known to diminish the
efficacy of neutrophilic phagocyte activity; this may allow for
some species to proliferate [48]. Increased growth of Gardnerella species can augment production of hemolysin, which may
further blunt any phagocytic protection that might otherwise
occur in the vaginal milieu [49].
GBS: perinatal group B streptococcal disease. Because of
the relative abundance of studies that deal with GBS, there are
sufficient fragments of information that one can use to infer
some aspects of intergenus bacterial regulation. Certain clues
lie in the demographics of the diseases caused by GBS. Although
GBS is a leading cause of perinatal and maternal postpartum
septicemia, the incidence of disease is grossly disproportional
to that of colonization [50, 51]. Depending on the use of special
media and the number of anatomical sites sampled, 14%25%
of women have GBS as a constituent of their vaginal flora. The
best statistics are those that correlate the incidence of perinatal
septicemia with material factors, including maternal antibody.
Prior to the implementation of protocols for avoidance of GBS
disease, the overall incidence of GBS perinatal septicemia was
1.23 cases per 1000 live births. The greater the quantity of
GBS present (i.e., the greater the density of colonization), the
greater the probability of disease [52, 53]. Maternal fever during
parturition is the factor associated with the highest incidence
of disease, followed by the presence of asymptomatic GBS bacteriuria in a gravida [54, 55]. A study of newborns with GBS
septicemia has demonstrated that isolates recovered from these
subjects have a greater ability to attach to epithelial cells than
do isolates from newborns without septicemia [56]. Although
some of the genetic requisites are known, the need for high
multiplicity of GBS has also been recognized. These observa-

REGULATION OF BACTERIAL FLORA

Bacterial interference is the term applied to in vivo situations in which indigenous microbial species regulate colonization by pioneering exogenous microorganisms. Bacterial interference can occur for a variety of reasons. These reasons may
include the production of antimicrobial substances by the interfering organism, the efficient use of some substrate in the
local environment, preemptive attachment to tissue sites, or a
more rapid rate of growth than that of competing organisms
[5963]. Quantitative relationships among bacterial species appear to be a key regulator of bacterial interference. The magnitude of the inhibitory effect may be the result of the potency
of the inhibitory substance and the number of producing
organisms.
In a study of consecutive cultures of vaginal flora specimens,
Carson et al. [58] introduced the term compatibility profiling
to describe the hypothesis that dominant regulatory bacteria
could be identified by virtue of their ability to emerge as the
sole isolate in samples in which numerical complexity would
ordinarily be observed. When this hypothesis was applied to
781 isolates, the only bacteria that achieved single-isolate status
were Lactobacillus species and G. vaginalis. Once these bacteria
were identified as sole isolates, analysis was extended to identify the co-isolate when only 2 bacteria were recovered. The
most prevalent of these bacteria were added to the initial key
bacteria isolated, and the process was repeated for cases in
which only 3 species of bacteria were recovered. The process
was again repeated with use of cultures when 4 species of bacteria were present. This iterative process of additive grouping
of bacteria established that bacteria such as coagulase-negative
staphylococci and the enterococci were compatible with both
e74 CID 2001:32 (15 February) Larsen and Monif

Lactobacillus species and G. vaginalis. By inference, those bacteria that were not present were presumed to be susceptible to
bacterial interference by the target bacteria or its subsequent
isolates.
Certainly, confirmation of this hypothesis will require additional in vitro evaluation. Lactobacillus species appeared to
be the major regulator of both G. vaginalis and selected anaerobic bacteria. Lactobacillus species were identified in 131
cultures of vaginal specimens. When Lactobacillus species were
present, G. vaginalis was a co-isolate in 7 cultures. In only 1
of these 7 cases were fewer than 5 isolates observed, including
the anaerobic bacteria present. In this study, inhibitory organisms appeared to include coagulase-negative staphylococci,
which appeared to suppress S. aureus and the group B streptococci, and other b-hemolytic streptococci.

THE ROLE OF THE LACTOBACILLI

Lactobacillus species are isolates that are commonly recovered
from cultures of vaginal specimens obtained from postpubertal
asymptomatic female patients. Quantitative studies have reported that vaginal washings contain 107 lactobacilli per gram
of secretion. The most common Lactobacilllus species include
L. acidophilus and L. fermentum; less common are L. plantarum,
L. brevis, L. jensenii, L. casei, L. delbrueckii, and L. salivarius.
More than 1 species may be present in an individual [64]. A
longitudinal study has shown variability in terms of species or
combinations of species over time [8].
In an in vitro study, Skarin and Sylwan [65] demonstrated
the ability of Lactobacillus species to inhibit the growth of several bacterial species, including G. vaginalis, Mobiluncus species,
Peptostreptococcus species, and Bacteroides species. They attributed this inhibition primarily to production of a low pH. Reid
et al. [66] suggested an alternate mechanism of control of the
bacterial flora by the lactobacilli. They found that cell wall
fragments of Lactobacillus species could block attachment of
bacterial uropathogens to uroepithelial cells. It is not clear
whether this observation might also apply to vaginal epithelial
cells or whether adherence of vaginal microorganisms to the
epithelium might be blocked by this mechanism. Colonization
of the introitus with Enterobacteriaceae species is a predisposing
factor for urinary tract infection in women. Raz and Stamm
[67] showed that estrogen therapy helped alleviate recurrences
of urinary tract infection in a cohort of women. Several lines
of evidence support a role for estrogen in increasing the density
of vaginal colonization by normal flora organisms [68, 69].
Special focus has been placed on the idea that hydrogen
peroxide production is a mechanism of bacterial antagonism
of the Lactobacillus species [7075]. Eschenbach et al. [72] advanced the concept that hydrogen peroxide, rather than pH, is
a prime regulatory feature of the lactobacilli. They detected

Downloaded from http://cid.oxfordjournals.org/ by guest on August 28, 2015

dealt with bacterial vulvovaginitis, Monif [19] identified 4 cases

of GBS vulvovaginitis for which both qualitative assessment
and some form of quantitative assessment were available, and
for which appropriate culture techniques had been employed.
In each case, GBS was present at a high multiplicity. In 3 of
the 4 cases, single co-isolates were identified: Escherichia coli in
2 cases and S. aureus in 1. GBS does not inhibit the growth of
either of these bacteria in vitro. This report indicates that when
the issue of bacterial density is effectively addressed, in vivo
observations tend to parallel those derived from in vitro
experiments.
The studies cited above suggest that a critical level of bacterial
replication must be achieved to support the disease-producing
capability of either an exogenous or an endogenous bacterial
strain within the female genital tract. Whether a bacterial strain
within the female genital tract attains the requisite population
level appears to be governed, to a significant degree, by the
microbes that are concomitantly present.

of lactobacilli. Zheng et al. [31] demonstrated that, although

hydrogen peroxide had little effect on the quantity of viable N.
gonorrhoeae in culture at neutral pH, the peroxide became more
effective at acidic pH. Although this result was obtained with
a catalase-negative organism, Larsen and White [30] showed a
similar result with the catalase-producing C. albicans. Perhaps
it is appropriate to conclude that probiosis in vivo is likely to
be multifactorial and that synergy between several factors exists.
THE CONCEPT OF PROBIOTICS
Recognition that one microbial species can inhibit a different
species of microbe has generated an interest in the exploitation
of this phenomenon for the benefit of the well-being of humans
or animals [78]. A term related to probiotics is prebiotics,
which refers to the feeding of substrates that promote the development of a benign microflora. For example, it is known
that consumption of fructo-oligosaccharides selects for the development of an intestinal flora dominated by Bifidobacterium
species [79].
Bifidobacterium is only one of several microbial genera that
have probiotic potential. Most proposed probiotics are grampositive bacteria, including enterococci, various Lactobacillus
species, Clostridium butyricum, and Bacillus species, in addition
to Bifidobacterium bifidus [80, 81]. In addition to these grampositive bacteria, the yeast Saccharomyces cerevisiae has been
used as a probiotic. Although promising probiotic approaches
to vaginal colonization are under consideration, the advice of
Gorbach [82] is as relevant to the vagina as it is to the gut,
The purported benefits for any probiotic must pass the highest
standards of scientific scrutiny before the claims can be accepted.
CONCLUSIONS
Studies have shown that the microbiological environment
may supersede the selected virulence of a given bacterial species
in the production of disease. The observations that suggest the
importance of environmental factors are intertwined with the
issue of microbial replication. For disease to occur, exogenous
or endogenous bacteria that possess pathogenic prerequisites
must attain replicative dominance. Their ability to do so is
potentially governed by inhibitory or synergistic interrelationships with other microbes. Although the lactobacilli are key
regulators when they are dominant in number, their ability to
maintain governance is influenced by other bacterial species
within the microflora of the genital tract.

References
1. Weinstein L. The bacterial flora of the human vagina. Yale J Biol Med
1938; 10:24760.

Flora of the Female Genital Tract CID 2001:32 (15 February) e75

Downloaded from http://cid.oxfordjournals.org/ by guest on August 28, 2015

Lactobacillus species in only 35% of women with BV; of those

women who were co-colonized with G. vaginalis and Lactobacillus species, only 11% had hydrogen peroxideproducing
strains.
Hillier et al. [75] demonstrated a significant correlation between the absence of hydrogen peroxideproducing lactobacilli
and vaginal colonization by G. vaginalis, Bacteroides species,
Peptostreptococcus species, and Mycoplasma hominis. There were
no significant differences between strains of Lactobacillus that
produced hydrogen peroxide and those that did not, with regard to concomitant isolation of enterococci, GBS or a-hemolytic streptococci, and catalase-positive bacteria, such as
diphtheroids, coagulase-negative staphylococci, and Enterobacteriaceae species. The prevalence of M. hominis or Ureaplasma
urealyticum was unaffected when cultures contained hydrogen
peroxidenegative Lactobacillus species or when no lactobacilli
were isolated from these cultures. When a very simple flora
exists, as it does in young adolescents, the lactobacilli are usually
dominant in number, and when only a single isolate is recovered, it is usually a Lactobacillus species. This is notable in view
of the close physical proximity of the vaginal introitus to the
perineum, with its abundant and complex flora.
Sexual activity, tampon use, childbirth, and various other
occurrences in the reproductive life of women are associated
with an increasing complexity of the flora, but one must ask
how Lactobacillus organisms are able to retain their dominant
status for long periods of time. Of equal importance, one must
ask how Lactobacillus species occasionally cease to be the dominant type of organism. Do other microorganisms have the
ability to emerge with the same dominance as lactobacilli in
some women? If so, what circumstances allow for the development of a flora that is not dominated by Lactobacillus species?
Inhibitory proteins have been isolated from strains of Lactobacillus acidophilus [76]. Holmberg and Hallander [62] documented the ability of Streptococcus sanguinus to inhibit the
growth of L. acidophilus, Lactobacillus fermentum, and Lactobacillus casei. Phonck [77] and Hillier et al. [75] reported that
streptococci may inhibit vaginal lactobacilli. Skarin and Sylwan
[65] used L. acidophilus and L. lactus to analyze bacterial inhibition on the predominant organisms cultured from women
with BV. Organisms such as G. vaginalis, Mobiluncus mulieris,
M. curtisii, Peptosteptococcus assacharolyticus, Peptostreptococcus
anaerobius, Bacteroides fragilis, and Peptococcus species (now
classified as Peptostreptococcus species) were inhibited by lactobacilli. In this study, the ability to acidify the medium was
better correlated with inhibition than was production of hydrogen peroxide. Skarin and Sylwan [65] found that L. acidophilus produced wider zones of inhibition on plate assays
and more lactic acid than did Lactobacillus lactis.
A number of reports have emphasized that production of
hydrogen peroxide is the key feature in the antimicrobial action

e76 CID 2001:32 (15 February) Larsen and Monif

28. Gujjar, P, Finucane M, Larsen, B. The effect of estradiol on Candida

albicans growth. Ann Clin Lab Sci 1997; 27:1516.
29. Larsen B, Markovetz AJ, Galask RP. Quantitative alterations of the
genital microflora of female rats in relation to the estrous cycle. J Infect
Dis 1976; 134:4869.
30. White S, Larsen B. Candida albicans morphogenesis is influenced by
estrogen. Cellular and Molecular Life Sciences 1997; 53:7449.
31. Zheng H-y, Alcorn TM, Cohen MD. Effects of H2O2 producing lactobacilli on Neisseria gonorrhoeae growth and catalase activity. J Infect
Dis 1994; 170:120915.
32. Girotti AW. Lipid hydroperoxide generation, turnover, and effector
action in biological systems. J Lipid Res 1998; 39:152942.
33. Monif GRG. Pathogenesis and classification of vulvovaginal candidiasis.
Am J Obstet Gynecol 1985; 152:9359.
34. Sobel JD, Chaim W. Vaginal microbiology of women with acute recurrent vulvovaginitis candidiasis. J Clin Microbiol 1996; 34:24979.
35. Savage DC. Microbial interference between indigenous yeast and lactobacilli in the rodent stomach. J Bacteriol 1969; 98:127883.
36. Monif GRG, Carson HJ. Female genital tract bacterial coisolates with
Candida albicans in patients without clinical vaginitis. Infect Dis Obstet
Gynecol 1998; 6:526.
37. Hipp SS, Lawton WD, Chen NC, Gaafar HA. Inhibition of Neisseria
gonorrhoeae by a factor produced by Candida albicans. Appl Microbiol
1974; 27:1926.
38. Shah D, Larsen B. Clinical isolates of yeast produce a gliotoxin-like
substance. Mycopathologia 1991; 116:203-208.
39. Shah D, Larsen B. Identity of a Candida albicans toxin and its production in vaginal secretions. Medical Science Research 1992; 20:3535.
40. McCormick WM, Hayes CH, Rosner B, et. al. Vaginal colonization
with Corynebacterium vaginale (Haemophilus vaginalis). J Infect Dis
1977; 136:7405.
41. Fredricsson B, Hagstrom B, Moller AK, et al. Bacterial vaginosis treated
with metronidazole: effects on the vaginal microbiology by a single
dose versus a five-day regimen. Zentralbl Gynakol 1986; 108:799.
42. Lugo-Miro VI, Green M, Masur L. Comparison of different metronidazole therapeutic regimens for bacterial vaginosis: a meta-analysis.
JAMA 1992; 268:92.
43. Thomason JL, Gelbart SM, Broekhuizen FF. Advances in the understanding of bacterial vaginosis. J Reprod Med 1989; 34:581.
44. Amsel R, Totten PA, Spiegel CA, et al. Nonspecific vaginitis: diagnostic
criteria and microbial and epidemiologic associations. Am J Med
1983; 74:1422.
45. Hillier SL. Diagnostic microbiology of bacterial vaginosis. Am J Obstet
Gynecol 1993; 169:4559.
46. Holmes KK, Chen KC, Lipinski CM, et al. Vaginal redox potential in
bacterial vaginosis (nonspecific vaginitis). J Infect Dis 1985; 152:37982.
47. Chen KC, Forsyth PS, Buchanan TM, et al. Amine content of vaginal
fluid from untreated and treated patients with nonspecific vaginitis. J
Clin Invest 1979; 63:82835.
48. Sturm AW. Chemotaxis inhibition by Gardnerella vaginalis and succinate producing vaginal anaerobes: composition of vaginal discharge
associated with G. vaginalis. Genitourin Med 1989; 65:10912.
49. Shubair M, Synder IS, Larsen B. Gardnerella vaginalis hemolysin. III.
Effects on human leukocytes. Immunol Infect Dis 1993; 3:14953.
50. Schuchat A, Wenger JD. Epidemiology of group B streptococcal disease:
risk factors, prevention strategies and vaccine development. Epidemiol
Rev 1994; 16:374402.
51. Prevention of perinatal group B streptococcal disease: a public health
perspective. MMWR Morb Mortal Wkly Rep 1996; 45(RR-7):124.
52. Ancona RJ, Ferrieri P, Williams PP. Maternal factors that enhance the
acquisition of group B streptococci by newborn infants. J Med Microbiol 1980; 13:27380.
53. Gerards LJ, Cats BP, Hougkamp-Korstanje JA. Early neonatal group B
streptococcal disease: degree of colonization as an important determinant. J Infect Dis 1985; 11:11924.
54. Persson K, Christensen KK, Christensen P, et al. Asymptomatic bac-

Downloaded from http://cid.oxfordjournals.org/ by guest on August 28, 2015

2. Gorbach SL, Menda KB, Thadepalli H, Keith L. Anaerobic microflora

of the cervix of healthy women. Am J Obstet Gynecol 1973; 117:10535.
3. Ohm JM, Galask RP. Bacterial flora of the cervix from 100 prehysterectomy patients. Am J Obstet Gynecol 1975; 122:6837.
4. Tashijian JH, Coulam CB, Washington JA. Vaginal flora in asymptomatic women. Mayo Clin Proc 1976; 51:55761.
5. Galask RP, Larsen B, Ohm MJ. Vaginal flora and its role in disease
entities. Clin Obstet Gynecol 1976; 19:6181.
6. Bartlett JG, Moon NE, Goldstein PR, et al. Cervical and vaginal bacterial flora: ecologic niches in the female lower genital tract. Am J
Obstet Gynecol 1978; 130:65861.
7. Hill GB. Anaerobic flora of the female genital tract. In: Lambe DW,
Genco RJ, Mayberry Carson KJ, eds. Anaerobic bacteria, selected topics.
New York: Plenum Press, 1980:3950.
8. Larsen B, Galask R. Vaginal microbial flora: composition and influence
of host physiology. Ann Intern Med 1982; 96:92630.
9. Brown WJ. Variations in the vaginal bacterial flora: a preliminary report. Ann Intern Med 1982; 96:9314.
10. Monif GR, Welkos SL, Baer H. Impact of diverging anaerobic technology on cul-de-sac isolates from patients with endometritis/salpingitis/peritonitis. Am J Obstet Gynecol 1982; 124:896900.
11. Bramley HM, Dixon RA, Jones BM. Haemophilus (Corynebacterium
vaginalis, Gardnerella vaginales) in a family planning clinical population. Br J Vener Dis 1981; 57:626.
12. Gopplerud CP, Ohm MJ, Galask RP. Anaerobic and aerobic flora of
the cervix during pregnancy and the puerperium. Am J Obstet Gynecol
1976; 126:858.
13. Bartlett JG, Onderdonk AB, Drude E, et al. Quantitative bacteriology
of the vaginal flora. J Infect Dis 1977; 136:2717.
14. Aly R, Britz MB, Maiibach HI. Quantitative microbiology of human
vulva. Br J Dermatol 1979; 101:4458.
15. Monif GRG, Jordan PA, Thompson JL, Baer E, Baer H. Quantitative
and qualitative effects of Betadine liquid on the aerobic and anaerobic
flora of the female genital tract. Am J Obstet Gynecol 1980; 137:4328.
16. Onderdonk AB, Zamarchi GR, Walsh A, Mellor RD, Munoz A, Kass
EH. Methods for quantitative and qualitative evaluation of vaginal
microflora during menstruation. Appl Environ Microbiol 1986; 51:
3339.
17. Levison ME, Corman LC, Carrington ER, et al. Quantitative microflora
of the vagina. Am J Obstet Gynecol 1977; 127:805.
18. Roy S, Sharma M, Ayyagari A, Malhotra S. A quantitative study of
bacterial vaginosis. Indian J Med Res 1994; 100:1726.
19. Monif, GRG. Semiquantitative bacterial observations with group B
streptococcal vulvovaginitis. Infect Dis Obstet Gynecol 1999; 7:2279.
20. Keith LG, England D, Bartizal F, Brown E, Fields C. Microbial flora
of the external os of the premenopausal cervix. Br J Vener Dis 1972;
48:516.
21. DeBoer JM, Plantema FHF. Ultrastructure of the in situ adherence of
Mobiluncus to vaginal epithelial cells. Can J Microbiol 1988; 34:75766.
22. Harris JW, Brown JH. The bacterial content of the vagina and uterus
on the fifth day of the normal puerperium. Bull Johns Hopkins Hosp
1928; 43:190200.
23. Thadepalli H, Savage EW Jr, Salem FA. Cyclic changes in cervical
microflora and their effect on infections following hysterectomy. Gynecol Obstet Invest 1982; 14:17683.
24. Pasteur L, Joubert JF. Charbon et septicemie. C R Soc Bio Paris 1877;85:
10117.
25. Smith T. Parasitism and disease. Princeton, NJ: Princeton University
Press, 1934:196.
26. Ramon AM, Porta A, Fonzi WA. Effect of environmental pH on morphological development of Candida albicans is mediated via the PacCrelated transcription factor encoded by PRR2. J Bacteriol 1999; 181:
752430.
27. Glover DD, Larsen B. Longitudinal investigation of Candida vaginitis
in pregnancy: role of superimposed antibiotic use. Obstet Gynecol
1998; 91:1158.

55.

56.

57.

58.

59.
60.
61.
62.

64.
65.

66.

67.

68.

69.

70.

71.
72.

73.

74.

75.

76.

77.
78.

79.
80.
81.

82.
83.

special reference to the Doderlein bacillus. Am J Obstet Gynecol

1939; 37:698703.
Wheter DM, Hirsch A, Mattrick ATR. Possible identity of lactobacillin with hydrogen peroxide produced by lactobacilli. Nature
1952; 170:6234.
Vincent JG, Veomett RC, Riley RF. Antibacterial activity associated with
Lactobacillus acidophilus. J Bacteriol 1959; 78:47784.
Eschenbach D, Davick PR, Williams BL, et al. Prevalence of hydrogen
peroxideproducing Lactobacillus species in normal women and
women with bacterial vaginosis. J Clin Microbiol 1989; 27:2516.
Redondo-Lopez V, Cook RL, Sobel JD. Emergence of lactobacilli in
the control and maintenance of the vaginal bacterial microflora. Rev
Infect Dis 1990; 12:85672.
Hillier SL, Krohn MA, Klebanoff ST, Eschenbach DA. The relationship
of hydrogen peroxide producing lactobacilli to bacterial vaginosis and
genital microflora in pregnant women. Obstet Gynecol 1992; 79:
36973.
Hillier SL, Krohn MA, Rabe LK, et al. The normal flora, H2O2 producing lactobacilli, and bacterial vaginosis in pregnant women. Clin
Infect Dis 1993; 16 (Suppl 4):S27381.
Mehta AM, Patel KA, Dave PJ. Purification and properties of the
inhibitory protein from Lactobacillus acidophilus AC1. Microbios
1983; 38:7381.
Phonck M. Streptococci antagonizing the vaginal lactobacillus. J Hung
Epidemol Microbiol 1962; 3:26770.
Reid G, Bruce AW, McGroaty JA, Cheng KJ, Costerton JW. Is there a
role for lactobacilli in prevention of urogenital and intestinal infections?
Clin Microbiol Rev 1990; 33544.
Roberfroid MB. Prebiotics and synbiotics: concepts and nutritional
properties. Br J Nutr 1998; 80:S197202.
Chimura T. [Ecological treatment of bacterial vaginosis and vaginitis
with Bio-three]. Jpn J Antibiot 1998; 51:75963.
Canganella F, Paganini S, Ovidi M, et al. A microbiology investigation
on probiotic pharmaceutical products used for human health. Microbiol Res 1997; 152:1719.
Gorbach SL. Probiotics and gastrointestinal health. Am J Gastroenterol
2000; 95(Suppl 1):S24.
Larsen B. In: Pastorck J, ed. Microbiology of the female genital tract.
In: Obstetric and gynecologic infectious disease. New York: Raven Press,
1994:1125.

Flora of the Female Genital Tract CID 2001:32 (15 February) e77

Downloaded from http://cid.oxfordjournals.org/ by guest on August 28, 2015

63.

teriuria during pregnancy with special reference to group B streptococci. Scand J Infect Dis 1985; 17:1959.
Persson K, Bjerre B, Elfstrom L, et al. Group B streptococci at delivery:
high count in urine increases risk for neonatal colonization. Scand J
Infect Dis 1986; 18:52531.
Helmig R, Halaburt JT, Uldbjert M, et al. Increased cell adherence of
group B streptococci from preterm infants with neonatal sepsis. Obstet
Gynecol 1990; 76:8257.
Chaisilwattana P, Monif GRG. In vitro ability of the group B streptococci to inhibit gram-positive and gram-variable constituents of the
bacterial flora of the female genital tract. Infect Dis Obstet Gynecol
1995; 3:917.
Carson HM, LaPoint PG, Monif GRG. Interrelationships within the
bacterial flora of the female genital tract. Infect Dis Obstet Gynecol
1997; 5:3059.
De Klerk HC, Cortzec JM. Antibiosis among lactobacilli. Nature
1961; 192:3401.
Brock TD, Peacher B, Pierson D. Survey of the bacteriocins of enterococci. J Bacteriol 1963; 86:7027.
Reves R. The bacteriocins. Bacteriol Rev 1965; 29:2545.
Holmberg K, Hallander HO. Interference between gram-positive microorganisms in dental plaque. J Dent Res 1972; 51:58895.
Tagg JR, Dajami AS, Wannamaker LW. Bacteriocin of gram-positive
bacteria. Bacteriol Rev 1976; 40:72256.
Rogosa M, Shape ME. Species differentiation of human vaginal lactobacilli. J Gen Microbiol 1960; 23:19720l.
Skarin A, Sylwan J. Vaginal lactobacilli inhibit growth of Gardnerella
vaginalis, Mobiluncus, and other bacterial species cultured from vaginal
content of women with bacterial vaginosis. Acta Pathol Microbiol Immunol Scand [B] 1986; 94:399403.
Reid G, Servin AL, Bruce AW, Busscher HJ. Adhesion of three Lactobacillus strains to human urinary and intestinal epithelial cells. Microbios 1993; 75:5765.
Raz R, Stamm WE. A controlled trial of intravaginal estriol in postmenopausal women with recurrent urinary tract infections. N Engl J
Med 1993; 329:7536.
Larsen B, Markovetz AJ, Galask RP. Quantitative alterations of the
genital microflora of female rats in relation to the estrous cycle. J Infect
Dis 1976; 134:4869.
Weinsten L, Howard JH. The effect of estrogenic hormone on the Hion concentration and the bacterial content of the human vagina, with